Facebook taught artificial intelligence to fool the systems, facial recognition in live broadcast

فيس بوك تدرب الذكاء الاصطناعي لخداع نظم التعرف على الوجه في البث المباشر

After the lawsuits received by Facebook as a result of its violation of the privacy of users by introducing them in the videos and images without permission through the systems of facial recognition; the company pipe the property list default, but it looks like they wouldn’t talk so to counter the support of the case against her.

As noted in the report, told VentureBeat that the research team of artificial intelligence in the company working on the development of new tools to modify its violations in this regard by artificial intelligence.

Where new tools unlike the mechanism of face recognition using machine learning to change the attributes of the fool the systems for face recognition in videos as well as videos of the live stream.

In no longer use of Facebook these tools to circumvent the systems of facial recognition is first in this area, where previously many companies follow the same method through the knowledge of the mechanism that is followed by the need to face and study its weaknesses.

And through the example of the attached model for this process we can see, after the small differences in facial features between my picture of Jennifer Lawrence before and after the application of the tools not to learn, but easy to identify as a possible American in both cases !

According to the report, the company does not intend to use this system in any product future no, maybe turn the researchers use this tool to protect users ‘ privacy and fight videos-counterfeiting deep that spread many applications recently, including the user data and affects them negatively.

Leave a Reply

Your email address will not be published. Required fields are marked *